Dynamic Spectrum Sharing (DSS) has emerged as one of the hot 5G discussion topics [see here on T-Mobile and Verizon]. This feature which allows simultaneous operation of LTE and 5G technologies has become a bellwether of product maturity and 5G leadership for equipment and silicon vendors and for service providers. But how important is DSS and is it worth all the attention?
What is DSS
DSS allows LTE and 5G to share the same frequency spectrum. The base station dynamically assigns resources to handsets operating on LTE or 5G technology. DSS allows operators to turn on 5G without expanding into new spectrum which makes it useful for spectrum constrained service providers. It is also attractive for service providers who want to claim 5G status quickly by leveraging their existing infrastructure and frequency spectrum.
How DSS Works
LTE and 5G radio waveforms share many similarities that allow for the two waveforms to coexist in the same spectrum. But there are also many differences that make 5G more efficient than LTE. One difference is how the pilot and synchronization signals work in LTE and 5G. A mobile handset uses the pilot signals to establish a common reference to synchronize with the network. Pilot and synchronization signals are critical to access the network and to communicate.
The pilot signals in LTE and 5G are different. LTE pilot signals occupy specific locations in the time-frequency continuum. On the other hand, 5G has more flexibility in assigning pilot signals. Combining LTE and 5G waveforms leads to collision of the 5G synchronization signals (called SSB: synchronization signal block) with those for LTE. To fix this problem, the 5G standard specified different techniques to avoid the collision of pilot signals. Two of these techniques are:
1. MBSFN technique (multicast-broadcast single frequency network): This solution occasionally blocks the 4G pilot signals and substitutes a MBSFN transmission for the blocked resources where it inserts the 5G synchronization signals.
2. Non-MBSFN technique: The system will use two LTE symbols in the frame and in their place will “squeeze” 4 5G synchronization symbols (SSB symbols) by shortening their duration.
Multiple factors would determine the choice between one technique or the other, including the frequency band of operation. This is important since DSS is better defined in some bands, such as n5 (850 MHz) and n66 (AWS). The 3GPP did not complete the definition of DSS in TDD mode.
Vendor Differentiation
DSS is a relatively complex feature because it involves scheduling across two different networks. The scheduler is the brain of the base station. How vendors implement the scheduler to prioritize access of 4G and 5G traffic to optimize performance is critical. Therefore, testing DSS in live networks under variable traffic conditions is important.
The granularity over which DSS operates is another aspect of differentiation. I have seen values ranging between 1 msec and 100 msec. Low granularity provides faster response to changing traffic requirements and gives better performance than high granularity.
The Disadvantage of DSS
Using DSS results in a loss of LTE and 5G capacity on the order of 10-20%. The amount of capacity loss depends on the mode and implementation technique of the equipment vendor. Hence, differentiation among vendors should focus on minimizing the capacity loss.
To put the capacity loss into perspective, overhead in LTE consumes about 20-25% of the channel capacity. 5G is more efficient and reduces the overhead to around 14% of channel capacity. Service providers seeking to maximize capacity will have to consider DSS in the context of their spectrum holdings and technology evolution strategy.
On the positive side, DSS allows the network to respond dynamically to demand from LTE and 5G users. Operators will have to weigh this benefit against the potential loss of capacity in making their deployment decision. Spectrum holdings will be key in the decision making process.
The Prospects for DSS
DSS targets the bands where LTE is already operational, which is mostly in the spectrum between 700 MHz – 2,700 MHz. Because DSS reduces capacity, it makes sense to use it in wide LTE channels such as 20 and 10 MHz.
DSS is useful in the early days of 5G roll outs when penetration of 5G devices is low and refarming of an LTE band is not feasible. 5G deployments will initially focus on the urban core leveraging mid-band spectrum. Therefore, DSS could come handy in low-density areas to provide continuous 5G coverage across all of the service provider market. In this case, the service provider needs to weigh the benefits against the cost of upgrading the equipment to 5G. Software upgrade of existing sites helps make the RoI more attractive.
Operators with large amounts of frequency spectrum would be interested to deploy DSS to claim 5G capability. This is especially the case in the US where access to mid-band spectrum is a couple of years away (primarily 290 MHz in C-Band spectrum which will be auctioned later this year). Both AT&T and Verizon fit this category. T-Mobile/Sprint is already rolling out 5G in 2.5 GHz spectrum in addition to considering 5G in 600 MHz, so DSS makes little sense for them.
Some operators claim that DSS in low bands will provide subscribers 5G service over a wider coverage area. This is correct, but the difference in performance between LTE and 5G in low frequency bands is not significant.
Operators with a small amount of spectrum cannot afford to lose capacity and will opt to wait until they acquire mid-band spectrum. Looking at the world’s map for mid-band spectrum allocation, a few service providers would benefit from DSS.
Concluding Remarks
Many of the international markets in Asia Pacific, Europe and other regions have allocated large amounts of mid-band spectrum for 5G. These markets will see little benefits from deploying DSS.
DSS is a feature for the early stage of 5G network deployments. It will be of little benefit in later stages. As 5G penetration increases, operators will refarm LTE spectrum for 5G. This is many years away: LTE is projected to peak in two years.